[unreadable] The integration of positron emission tomography (PET) and X-ray computed tomography (CT) into a single instrument allows acquisition of functional and structural images that are spatially registered. This provides the opportunity to use the CT data from a combined scan for attenuation correction of the PET data. Algorithms for transforming the CT image, which represents average attenuation at CT energies, to a 511 keV attenuation map generally produce a high quality attenuation correction. However, several types of inaccuracies in CT-based attenuation correction have been observed that result in inaccurate corrected emission images. One cause of deficiencies is the presence of IV or oral contrast agent administered to enhance the CT scan. Regions containing contrast agent often result in significantly overestimated attenuation values in the 511 keV map. Focal contrast artifacts in the resulting attenuation corrected emission images have been observed that resemble diseased tissue. Another type of artifact can occur in corrected emission images due to differences in respiration phase at the time of CT and PET acquisition. Generally, a PET acquisition will be averaged over many respiration cycles while the CT will be acquired over only part of a cycle or with a breath hold. The hypothesis of this research program is that the measured emission sinograms contain a sufficient amount of attenuation information to allow correction of some of the observed artifacts. The goal of the work is to develop, implement, and test methods for recovering limited defective regions of an otherwise ideal attenuation map by using this information. There exist several methods that have been used with only limited success for the general problem of extracting a complete attenuation map from emission data alone and which will serve as the starting point for this project. An important point is that the current research is aimed at recovering only limited regions of the attenuation map. This is a far simpler problem than the general case. The results of the proposed research will provide basic tools and experience for developing a comprehensive strategy for optimizing the use of PET/CT data. [unreadable] [unreadable]